Concerns about a Mains transformer's rating.

[bikerhifinut]

Hi gents.
Mulling over the Mulard 5-20 circuit as I have a set of transformers and chassis that are alleged to be suitable for building an EL34 amplifier such as the 5-20.
However, although the mains transformer has a secondary of 410-0-410V as per Mullards circuit, it's only rated at 150mA. Now, according to the Mullard book of circuits, the amplifier is reckoned to draw around 145mA as designed.
Am I sailing a bit too close to the wind here and should I consider different transformers? The heater windings are 5V@2A for a GZ34, and 6.3V@5A.
The output transformers are Danbury DB910 and are well up to the power output of a pair of EL34, the data I have with them says that they are good up to 35W from a pair of KT88 (they are suitable for EL34/KT66/Kt88 according to the data sheet).

I have a feeling the mains transformer should really be able to deliver closer to the 180mA specified by Mullard, even though that was to take into account supplying a separate preamp and/or tuner.


Am I ok or should I be worried?

Andy.

[G8HQP Dave]

Is the 150mA rating an AC RMS rating or a 'DC out from PSU' rating? If the former (most likely) then 150mA RMS from the transformer is not enough to feed 145mA DC. Even 180mA is sailing close to the wind.

The 5-20 runs the output almost in Class A so you need a transformer VA rating about 4 or 5 times the audio output power.

[kalee20]

I would not be worried, though I would ensure good ventilation.

Transformers should generally give their rated voltages at their rated currents. Significantly under-loading them could result in voltage being higher than nominal,which is not quite what you want, especially for valve heaters.

So if you are pulling 145mA, then a 150mA transformer should be ideal.

(Strictly, there are issues such as RMS-to-DC current ratios etc,but a transformer such as this would generally be specified in terms of DC current after rectification anyway).

[G6Tanuki]

In a normal domestic environment and with a typical usage duty-cycle I'd not worry. If you are worried, try strapping a thermometer to the txfmr core and see how hot it gets after an hour or so.

If you were going to be using it 24/7/365, and any failure was going to result in airplanes crashing out of the sky or ships sinking with all hands in mid-Atlantic - then I'd be looking more closely at the ratings relating to permitted temperature-rise and other such.

With a modern transformer if it passes the 100C test (no sizzle from a wet finger) it will be alright. And modern stuff is rated for quite a high ambient temperature, if the it is in a cooler environment you can squeeze out a bit more.

[daviddeakin]

I also would prefer to adhere to Mullard's original spec of 180mA.

If you use a silicon rectifier instead of the GZ34 the absence of heater demand will at least buy you some extra VA for the HT winding.

[broadgage]

The voltage rating suggests a centre tapped HT winding with the output rectified by a double diode.

The DC output current is therefore supplied half from each side of the secondary, and even allowing for losses, the DC output can exceed the RMS AC rating of the transformer.

Might be easier reckoned in watts.
Transformer is rated at 820 volts at 150ma, or 123 watts.
With a perfect rectifier and no smoothing that would give 410 volts at 300ma output, still 123 watts.
In practice voltage will be dropped in the rectifier and choke, but the output voltage will also be increased by the capacitors charging up to nearly the peak of the AC volts.
All this results in an output appreciably less than 123 watts, but I would still expect the DC current available to be greater than the AC rating of the transformer.

Or put another way, what if you used instead a 410 volt, 300ma transformer and a 4 diode full wave bridge rectifier, what DC current would you expect ?
More than 150ma I suggest, but less than 300ma.

[G8HQP Dave]

Originally Posted by broadgage

Transformer is rated at 820 volts at 150ma, or 123 watts.
With a perfect rectifier and no smoothing that would give 410 volts at 300ma output, still 123 watts.

This mixes up so many issues of RMS vs average in both current and voltage!

Transformers are usually specified in terms of RMS voltage output. A centre-tapped 820V winding would give 410V RMS (less diode drops). Unsmoothed this gives about 369V (=0.9x410) average DC (but with huge ripple). Add choke input smoothing and you get smooth 369V DC. Use capacitor input smoothing and you instead get 580V DC (=1.414x410). A valve rectifier will drop some voltage, though. What you are most unlikely to get from a 410V RMS winding is 410V DC output.

Transformers are now usually specified in terms of AC RMS current output. The RMS current drawn by a capacitor input supply can be 2 or 3 times the DC output current. The reasons some people seem to be able to do better than this are either Class B duty cycle with real music (so full DC current only drawn for a tiny fraction of operation time), or equipment only used for a limited time so the transformer does not have enough time to build up a dangerous temperature. Note that it is not core temperature which matters but winding temperature. Some older transformers may be specified in terms of DC current output from a typical PSU, and valve rectifiers lead to a smaller ratio of DC vs RMS current as they are so lossy themselves.

Note that a centre-tapped transformer used with a pair of rectifiers (solid-state or a single dual-anode valve) is less efficient than a single winding used with a bridge. This is because you need twice the length of wire so it needs to have half the cross-sectional area, so twice the resistance so only 70% of the RMS current rating. The only reason such transformers were used in the valve era is that valve bridges required extra envelopes and heater windings, so the simpler common cathode circuit was adopted.

[bikerhifinut]

Originally Posted by broadgage

More than 150ma I suggest, but less than 300ma.

Hi,
Yes that's what conclusion I came to in the end.
I realised it's a 120VA transformer on the HT side, so the figure you calculated (and me also) came more or less on the button.
I did the RMS calculation and at peak volts it should still give in the region of 200mA, so well within spec. I basically divided the 300mA by root2, I hope i got that bit of theory correct.
I'm using a GZ34, if there's room for a choke I shall put one in but it looks like its getting a smoothing resistor instead due to lack of physical space on the chassis. I am basically using the HJ Leak circuit from the TL25+ but adapted for octal valves in the preamp and phase splitter as the chassis has octal holes punched in. I have a set of ECC81 in case, if its possible to make a B9A to octal adaptor for them, but my homework so far suggests I can use the 6SL7 that I also have in stock.
That's the main winter project, I should have my PCL86 prototype finished by the end of the weekend with luck.

A.

[jimmc101]

There's a useful crib sheet from Hammond Manufacturing... Design Guide for Rectifier Use

Nice and simple for the Full Wave CT capacitor input filter. Idc = Iac rms

Jim

[G8HQP Dave]

I would not attach too much importance to that page. It mentions Vavg DC for capacitor input circuits where Vavg DC has no meaning.

[G6Tanuki]

If you're really worried:: wire some suitable load-resistors to the secondaries and shove 240V up the primary for an hour or so then see how hot it gets.

The "spit-test" as mentioned above is a good assessment.

Your amplifier's not a safety/life-critical application so don't get uptight about it: you're more likely to die from the elevated blood-pressure worrying about these things than from the transformer exploding!

[ex 2 Base]

jimmc 101 are pages 1-10 of the Hammond booklet interesting reading. I went to there web site but was unable to obtain them.Ted

[dominicbeesley]

The thing I learned with my first amplifier project was that transformers can be run close to their limit - but then they need lots of ventilation! My first PSU's were in enclosed boxes and quickly got too hot to touch, an afternoon drilling lots of holes has solved it and ten years later they've not gone up in smoke.

[jimmc101]

Originally Posted by ex 2 Base

jimmc 101 are pages 1-10 of the Hammond booklet interesting reading.

It's just their transformer catalogue transformer catalogue (! 10MB)

As G8HQP Dave has pointed out there are errors on that page V(Avg)D.C. is incorrect.

For an unsmoothed full wave rectified sine wave (theoretical, no losses)

V(Peak )D.C. = 1.414 V(rms)A.C., V(Avg)D.C. = 0.9V(rms)A.C.

Add a reservoir capacitor and

VD.C. = 1.414 V(rms)A.C.


A much fuller description is available in 'Mullard Technical Communications' Volume 7 Number 68 Jan 1964 'Power Rectification with Silicon Diodes' but I can't find a copy on the net.

Jim

[GMB]

Originally Posted by broadgage

Transformer is rated at 820 volts at 150ma, or 123 watts.

Well is it?
I have often wondered what exactly was meant when a transformer with a centre-tapped winding rather obviously meant for a double-diode rectifier say XXXmA as the current rating. Does it mean that each winding could deliver that a/c current or is it assuming that the pair deliver that current with alternate cycles coming from each winding?

These days there is usually a VA rating that makes it clearer.

[daviddeakin]

'Valve' transformers like this one nearly always quote the DC load current under normal operating conditions. They don't quote the RMS secondary current or VA rating like ordinary transformers. The OP's transformer is certainly rated for 150mA DC with a standard capacitor-input rectifier (power factor about 0.7), which is not really enough for a 5-20. Omitting the GZ34 would relax matters.

And as others have noted, that Hammond cheat cheat is not to be trusted